A (long) Comment on Scott Alexander’s Review of “Unlearn Your Pain”

My son Ben texted me the other day that his favorite blogger, Scott Alexander, had published a post reviewing Howard Shubiner’s book Unlearn Your Pain.  Despite being a psychiatry resident, Scott manages to be a prolific writer in areas of interest which appear to be TNTC.  Ben knows him through Effective Altruism, and sent me the review because it mentions ISTDP (Intensive Short-Term Dynamic Psychotherapy), the kind of therapy that I do. I had read some of Scott’s other posts with interest, always wishing I had more time to keep up.

I enjoyed the review, and sat down to email a response to Scott, since my thoughts seemed a little long to post as a comment. Scott asked me to post them as a comment anyway, at which point I discovered that his blogging platform agreed with my initial gut feeling.  So I am posting them here, and will post a link to this post in the comment section of his review.

So here is a somewhat edited version of the email that I sent him.

Hi Scott,

My son Ben Kuhn texted me the other day about this post. At his urging I’d read some of your previous posts; I’ve enjoyed and wish I had time to read more. For this one, as a psychiatrist who practices and teaches Davanloo’s Intensive Short-Term Dynamic Psychotherapy (ISTDP), I rushed-right over when he texted me!

Howard Schubiner is an exemplary human being; he has referred a number of patients to me and he has spent a great deal of time corresponding or speaking with almost all of them. Multiply this by a nation of 320 million and I find the implications quite stunning.

Unfortunately, he may have set up a straw man in terms of theory of psychosomatic pain. Having worked for a while after residency as a CL psychiatrist, my overall impression is that psychosomatic ailments are a vast swamp, and it is very easy to get sucked down into the mire, especially if you are theorizing about their cause. And, in general, early medical theories are fairly laughable in retrospect, if not at the time. As an example, the entire premise of the DSM is fairly laughable, it’s just that it is (in some ways) better than what we had before.

So here is your summary of Howard’s theory:

He thinks there may have been some original minor injury, of the sort that most people would get over in a couple of weeks. This causes the nerves to “sensitize” – ie the brain is primed to think about and remember this form of pain.

Here is your alternative:

But I’m not convinced by its more sweeping claims that physical injuries play little-to-no role in chronic pain. Along with Schubiner’s talk of nerve sensitization, one can imagine a scenario in which alternatively apparently-healed physical injuries may leave very small irritations on local nerves, and that the degree of irritation a nerve is able to bear without giving you chronic pain is related to your general neural-non-messed-upness and stress level. In such a scenario psychological factors might play a role in gating the pain, or in tensing or releasing muscles around the pain, but would not entirely explain it.

In his version, the locus of the problem is a “sensitized” nerve connected to a healed injury. In your version, it is a nerve hooked to an injury that is healed enough that people with a non-messed-up nervous system would not perceive it as noxious. At our level of ability to understand, observe, and intervene in these phenomena, I’m not able to detect any practical difference between these two points of view, and so discussions of this kind often turn out to be a waste of breath. A nicer way to say that is that (a) at our current level of measurement, these are not empirically verifiable or falsifiable claims, and (b) even if you could answer this question one way or the other (and perhaps both are true in different cases), it is not clear whether that would have any bearing at all on the question of whether any given treatment is effective or not. For me, and I suspect for you as well, the bottom line is (a) does a particular treatment relieve human suffering, and if so (b) is the time, money, sweat and ancillary suffering need to do the treatment worth it for the benefit it has? I don’t think that we’re at a point where theories of this kind can illuminate these questions, beyond the first step of making it plausible that a psychological treatment could help a “physical” problem like pain (though why we are comfortable with the conclusion that pain is a physical problem when there is no physically observable injury is an interesting question). Your hypothesis and Howard’s both fit the bill for that first step.

Oops, I honestly don’t mean to be getting into bashing you here, so I am going to put my discussion of placebo toward the end.

Second: As much as I like and admire Howard, I would not describe what he does as ISTDP; it is ISTDP-friendly, but all of the “silly-sounding” things you describe in your review are things that I’ve never seen an ISTDP therapist do (and since we videotape our therapy and every presentation involves showing video, this statement actually means something). They sound kind of silly to me, too, though I’ve seen sillier-sounding things, such as EMDR, really help people so I do admire your restraint and I think it is a wise approach. I don’t think that much of what I do looks silly, but some of it may look kind of crazy… but it keeps helping people, so I keep doing it.

I think it is more accurate to say that Howard is in the John Sarno tradition than to say he is doing ISTDP. Sarno also had close links with ISTDP; in fact I’ve been told that a psychologist he worked with trained with Davanloo, so that it is possible that much of his thinking originated with Davanloo, but there is a lot of thinking that originated with Davanloo that then diffused out, but is not ISTDP. For example, I am told that Jeffrey Young, the originator of CBT’s “schema” therapy, attended Davanloo workshops.

One of my main discomforts with Sarno is that he tells people “you have to believe that your pain is caused by TMS” (his name for what we’re talking about, Tension Myositis Syndrome). This is not a tenet of ISTDP; in fact, I welcome patients’ skepticism and encourage them to take an empirical, observational, experimental stance toward their interior life. All that is necessary is a willingness to try it out and see whether it’s helpful.

Next: “childhood trauma.” ISTDP says that “neurotic” psychopathology is based on “attachment trauma,” i.e., some kind of trauma to early attachment bonds. So if you find Howard’s claims uncomfortably broad, this is probably even more uncomfortably broad. So that we don’t get caught up in hair-splitting let’s replace “is based on” to “has a very significant component based on.” Attachment trauma includes everything that comes to mind when you say “childhood trauma”: physical abuse, sexual abuse, early parental loss. It also includes covert forms of abuse or neglect. It is possible that it is so broad as to be universal. It is undoubtedly true that there are events which are traumatic to one child but not to another, based on factors that we do not necessarily understand but which may well include a temperamental/genetic component. It is conceivable that a prospective study based on observable putatively pathogenic events would show only a fairly weak correlation with subsequent psychopathology. The point of this theory is not really to predict who will become ill; it is to indicate the kind of therapeutic work that helps people get better. This also does not mean that “therapy that focuses on childhood trauma will help people get better,” or even that “therapy that focuses on attachment trauma will help people get better.” ISTDP has a very specific way of focusing attachment trauma. In addition, it is easy to do ISTDP poorly, and not so easy to do it really well. This means that, while the claim that “Well-conducted ISTDP helps people get better” is potentially verifiable, it is not easy to verify. So I am aware that there is a lot of potential refuge for scoundrels here. However, in terms of your post here, the relevant point is that studies showing a weak link between “childhood trauma” and psychosomatic complaints aren’t necessarily convincing evidence against the ISTDP point of view.

Listen, I could go on and on, and there is some danger that I would (actually, reading this over, I see that I have). But the point is, chronic back pain is an enormous problem from a public health perspective. How much of it is amenable to psychological treatment that would have a strong result (i.e. “it solved my problem [completely or almost completely],” rather than “yeah, I think it probably helped some”). 20%? 90%? 99%? 50%? 10%? I don’t think we have much idea. Even if it’s 10%, that would be very significant from a public health perspective. Chronic back pain is one piece of chronic pain, albeit a large one. Chronic pain is one piece of medically unexplained symptoms. Medically unexplained symptoms comprise a stunningly large percentage of human suffering, doctor visits, and chew up an enormous chunk of our vast medical budget. If you add to that the fact that our treatments for issues that ARE regarded as psychological leave a lot to be desired, we are talking about potentially addressing an enormous amount of human suffering and healthcare expenditure.

One of my main authorities on ISTDP is Allan Abbass, a psychiatrist at Dalhousie University in Halifax, Nova Scotia (www.istdp.ca). After his initial training in family and emergency medicine, he trained extensively with Davanloo. He’s gone on to do quite a lot of research on ISTDP and has surveyed the rest of the research for Cochrane reviews among others. If at some point you want to take a deep dive into the research base you might wait a bit because they have an RCT of ISTDP for treatment-resistant depression that hasn’t yet been published but is on the way. Here is a link to an article describing ISTDP and presenting some of the research. I do hasten to add that ISTDP theory is not particularly attractive, as theory goes. As a former mathematician, I am a big fan of elegance. Sadly for elegance, the more closely I follow the theory, the better my patients get.

Some of Allan’s results are quite stunning, e.g. an ED-based intervention for Medically Unexplained symptoms with an average of 3.8 sessions showing something a > 50% reduction in ER utilization in the year after intervention compared to the year before, much greater than matched controls. Symptom relief in ISTDP can be very rapid, though for various reasons musculoskeletal pain can be less rapid than other somatic symptom disorders like migraines, IBS, nonepileptic seizures, etc. The article can be found here.

In terms of my own history, I was positively disposed toward dynamic psychotherapy when I started training, and I trained in a dynamically-oriented program (Mass. Mental Health Center, which then became part of the Harvard Longwood residency program). Over the course of my training I became gradually disillusioned (I won’t say disenchanted because I wasn’t really enchanted to start with) because it seemed to be a welter of competing speculation with no way of discerning the validity of any of it. My experience with ISTDP has been different from that, because you see so much unfolding in front of you in a very explicit, experiential way. It is fair to say I am enchanted with ISTDP, with most all of the positive and negative implications of that word.

For various complicated reasons, the ISTDP world is more fragmented than it needs to be and there has been quite a bit of terminological confusion. I wrote a book called “Intensive Short-Term Dynamic Psychotherapy: A Reference” which is something like a cross between a glossary and a series of blog entries. It is not designed to be read cover-to-cover, though some people have and said that they enjoyed the experience. But if you want to learn more about ISTDP from a book, I would say that the place to start is Allan’s book, “Reaching through Resistance.” Both are available on Amazon, but I also need to say: ISTDP can’t be learned from a book, you definitely need to see video of the therapy, and you definitely need supervision. Allan also offers a 3-day “immersion” in ISTDP every fall.

Here’s one thing I enjoy about ISTDP. You undoubtedly remember from med school hearing about “symptoms and signs” of disease, and learning the distinction very early on. You are obviously also aware that in psychiatry we say a lot about symptoms and zero about signs. (For your non-medical readers: symptoms are things that patients complain about, signs are things that physicians may observe but that patients do not complain about.) Well, ISTDP has honest-to-God signs, in terms of what Davanloo calls the “discharge pathways of unconscious anxiety,” and this is one of the main guides to the therapist in doing the therapy.

Oops, I see that I said I’d say something about placebo. You clearly know the placebo literature much better than I do, though that isn’t hard. But is seems to me that when we talk about placebo, we are talking about the mind’s ability to heal the mind and body. What could be better than that? Yet we talk about it as though it is noise, rather than perhaps the most valuable of signals. So I agree that we should be doing everything we can to develop and augment the “placebo” effect. If putting on robes and chanting makes it work better, let’s put on robes and chant. If I ever do learn the placebo literature better, the title of the article I want to write is, “Is it Ethical to Withhold Placebo Treatment?” The point I want to make here, though, is that while it is (mostly) possible to make a viable distinction between medical interventions and placebo, I’m not at all sure that you can make a viable distinction between psychotherapeutic interventions and placebo. Oops, we are at the point of more than “more than enough” so I’ll stop… I did have a question about your remarks on behavioral genetics but I’ll have to save that for another day…

4 Replies to “A (long) Comment on Scott Alexander’s Review of “Unlearn Your Pain””

  1. An important question to answer about the placebo effect is whether a placebo effect can be long lasting, or whether it only provides temporary relief until the patient forgets or adapts.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: